The bottom of my prints warp/curve upwards, most often at the corners. This is a very slight curve, only about 1-2 mm.

  • I print with a raft all the time.
  • I don't have a heated bed.
  • I print with PLA at 199 °C (390 °F) with a print speed of approximately 40 mm/s
  • I have a Sindoh 3DWOX DP201.

What slicer settings might be the cause of this phenomena? Or could it be 3D printer settings?


2 Answers 2


The first thing to understand is what causes warping. Warping is caused by the thermal contraction of the plastic when it cools down.

Simplifying things a fair bit, you can visualise the process like this:

  1. hot, expanded plastic gets deposited on cooler, shrunk layers,
  2. when the hot plastic cools down, it shrinks and pulls the upper part of the layer below inwards
  3. at this point, the layer below has a differential in the compression between its upper and lower parts, and curls up
  4. the problem is exacerbated at the very first layer (the one touching the bed) as this is "locked" to a rigid body (the bed) and cannot shrink, while subsequent layers are only attached to the somewhat flexible plastic beneath, and thus can contract.

diagram of curling plastic

Also notice that the larger the part being printed, the stronger is the force trying to curl-up your print.

Once one understands all of this, then it is possible to appreciate the many ways the problem can be mitigated.

Here are the common ones:


This translates in smaller tensions and thus less force "pulling up" the corners of your print. Historically, 3D printing started with ABS because this material was one of the very few, relatively safe ones to source. Nowadays there are materials like PETG which have similar mechanical properties to ABS but are much easier and forgiving to print with, so - unless you need ABS for some very specific reason (e.g.: acetone smoothing) consider never printing with it.


Concretely, this means lowering the "gap" between the ~200°C of the nozzle and the ~20°C of room temperature by using a heated bed and - possibly - an enclosure.

The heated bed not only drastically diminish the shrinkage of the first layer, but because heat radiates, and hot air goes upwards, the entire bottom of the print has shrinkage mitigated.

An enclosure just increase the benefit of the heating bed, by reflecting IR radiation back towards the print and preventing hot air to escape. A heated enclosure just improve things even further.

Some slicers offer a "shroud" option, that encloses the entire print in an enclosed, sacrificial structure, that tries to emulate the benefits of a proper printer enclosure.


That is the "brutal force" approach: if you face a strong "curl up" force, oppose it with a strong "anchor down" one.

The increase in adhesion can be achieved in a number of ways:

  • Lower print speed (more time for the molten plastic to "bond")
  • Overextrusion (more pressure, more material)
  • Disabling cooling fan (more progressive cooling, more time to "bond")
  • Using a brim (more contact surface between print and bed)
  • Using "ad hoc" material on the bed (PVA glue for PLA, ABS sludge for ABS, kapton tape, hair spray, blue tape, etc...)


This is typically achieved during design. Designing is a vast field and it would be impossible to cover all the possible mitigating strategies one could use, but here are some of the most common ones:

  • Prefer assembling smaller parts over printing huge ones. This is self explanatory really, as the curling force increases with the amount of material "pulling", the least material one has, the less force one gets.

  • Make relief holes above the first layers in long structures. This will essentially "break" the build-up of tension in the layer, creating many points with a little "curling up force" rather than two with a huge one. Something along the lines of this, for example:

I beam with relief holes

  • Avoid extensive overhangs close to the bottom of the print (this is because otherwise you will have considerably more material "pulling up" than you will have "anchoring down". Here is an example of what not to do (to be fair: this was specifically taken from a bed adhesion/warping test).

warping test

Of course all of the above strategies can/should be combined, when possible. Even if not warped, a part with a lot of internal tension will perform less predictably and possibly worse than a part where such tensions are lower.


If you have a heat bed, heat it up accordingly (for example for PLA 50 °C first layers, 40 °C then can be a good starting point).

If the first layer isn't close enough, then warping can happen (Happened to me when moving from 0.3 mm layers to 0.1 mm).

If you are under extruding / have dirt in the system (the heat chamber) so that extrusion is uneven it can make warps.

And as always, you can try to print slower, it helps not always but often.

Please also do tell what temperatures (heat bed & nozzle) you are using and what material.

  • $\begingroup$ You do not want to drop bed temperature mid-print just because you're past the part that will warp. Unless you've mashed the filament into the bed with a nozzle that's way too low, turning the bed temp down will severely break adhesion due to different thermal expansion of the print from the bed material/adhesives. If starting with 50, stay with 50 til the print is ready to remove from the bed. $\endgroup$ Dec 7, 2020 at 18:22
  • $\begingroup$ The higher temperature is only for the first layer, that's not a big difference to say, the plastic coming out of the nozzle. There will always be forces trying to warp so you have to contain them in some sort of way (object sticks well to the plate, forces gets absorben in the object as it cools slowly...). Just to add, I'm usually going for much higer bed temps nowadays, up to 90° for the first layer for some complicated plastics and then -10°. Would love to have a better control of the temp through the print though... $\endgroup$
    – Valmond
    Dec 8, 2020 at 14:01

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .